Customer 360 is a single, complete view of each customer, built by bringing together what every system in the business knows about them: who they are, what they have bought, what they have asked for, what they owe, and what has been promised to them. The aim is that anyone, or any AI agent, dealing with that customer sees the same full picture.
The idea is not new. What has changed is who reads the view. A person can cope with a customer 360 that is slightly out of date or missing a field. An AI agent acting on it cannot tell, and will act anyway.
What a customer 360 includes
- Identity: names, contacts, legal entities and the links between them, matched across systems.
- Relationship: accounts, contracts, products held, pricing and terms.
- Transactions: orders, invoices, payments and returns.
- Interactions: support tickets, calls, emails, meetings and website activity.
- Risk and compliance: credit status, verification checks, consents and restrictions.
- Commitments: open quotes, promised call-backs, agreed exceptions and discounts.
Most projects deliver the first four. The last two are the ones that prevent mistakes.
How a customer 360 works
- Integrate. Bring data in from the CRM, billing, support, product and marketing systems.
- Unify. Match records that belong to the same customer (identity resolution), settle conflicts, and build one profile.
- Activate. Make the profile usable where work happens: in the service desk, in sales tools, in reports, and now by AI agents.
Most of the effort, and most of the failure, is in the second step.
Customer 360 vs CRM vs CDP vs MDM
| What it is | What it is for | What it lacks alone | |
|---|---|---|---|
| CRM | System for managing sales and service relationships | Pipeline, accounts, cases | Billing, product usage and data held in other systems |
| CDP | Platform that collects and unifies customer behaviour data | Marketing segments and personalisation | Contracts, finance, risk and commitments |
| MDM | Discipline and tools for one agreed master record | Clean, matched identity | Transactions and interactions |
| Customer 360 | The outcome: one complete view of the customer | Every team and agent working from the same picture | Nothing by definition, but it depends on the three above |
Why customer 360 projects stall
- No agreed identity. The CRM, billing system and support desk each have their own customer record, and nobody owns the match between them.
- Conflicting values. Two systems hold different addresses or credit limits and there is no rule for which one wins.
- Stale copies. The 360 is a nightly copy, so it is wrong for anything that happened today.
- Data without meaning. The view shows a 15% discount but not that it was a one-time exception.
- Built for dashboards. The result is a screen for people to look at, with no way for another system or an agent to use it safely.
Customer 360 for AI agents
An AI agent that answers a customer, adjusts an order or approves a refund needs more than a profile. It needs to know:
- Which source is authoritative for each fact, so it does not act on the wrong address or limit.
- How fresh each fact is, so it can refuse to act on old data.
- What rules apply to this customer: consent, region, contract terms, approval limits.
- What has been decided before, including exceptions, so it stays consistent with past handling.
- What it is allowed to do, checked before it acts, with a record kept afterwards.
This turns customer 360 from a reporting project into a decision project. The question is no longer "can we see everything about the customer?" but "can an agent act correctly on what it sees, and can we show why it did?"
Customer 360 use cases by industry
- Banking and financial services: onboarding and KYC, fraud and AML review, suitability checks, relationship pricing.
- Retail: returns and refunds, loyalty, consistent service across store and online.
- Manufacturing: order status, warranty and service history, account credit and terms.
- Telecom: billing disputes, plan changes, retention offers.
- Travel and hospitality: disruption handling, guest preferences, loyalty recognition.
Are you ready? Six questions
- Can you match one customer across your main systems today?
- For each key field, do you know which system is the source of truth?
- How old is the data when someone uses it?
- Are consents and restrictions stored with the profile?
- Are exceptions and promises recorded anywhere a system can read?
- Can you show afterwards what information a decision was based on?
Three or more "no" answers means the work starts with identity and ownership, not with a new tool.
How to approach it
- Start from decisions, not data. List the customer decisions you want made faster or automated, such as refunds, renewals, credit holds or onboarding, and bring together only what those need.
- Settle identity first. Agree one way to match a customer across systems and one owner for it.
- Name the winner. For every important field, record which system is the source of truth.
- Read live where it matters. Balances, order status and consent should come from the system of record at the time of the decision.
- Attach the rules. Keep policies and limits alongside the data, with versions.
- Record every action. Each decision made on the view, by a person or an agent, should leave a trace of what was used and why.
Where ElixirData fits
ElixirData's Context OS assembles the customer context an agent needs at the moment of a decision, from governed sources, with the applicable policy attached, and records the result as a Decision Trace. It works with the CRM, ERP and data platforms already in place; it does not replace them. The record of past decisions it builds is a context graph.
Frequently asked questions
Is customer 360 the same as a CDP?
A customer data platform is one tool often used to build a customer 360, mostly for marketing. Customer 360 is the outcome: one complete view, however it is built.
Is customer 360 the same as master data management?
Master data management solves the identity part: one agreed record per customer. A customer 360 adds transactions, interactions and commitments on top.
Is customer 360 a Salesforce product?
Salesforce uses "Customer 360" as a name for its product suite. The term is also used generally, as here, for a complete view of the customer on any technology.
Do we need a data warehouse first?
A warehouse helps with history and analysis. For decisions that depend on current facts, the view also needs live reads from the systems of record.
How long does it take?
It depends on how many systems are involved and how clean identity is. Starting with one decision and a few sources shows value sooner than attempting every source at once.
Want to see governed customer context on your own systems? Request a demo.
